st: How to assign cumulated frequencies to observations

[Timm Fulge <t.fulge@uni-bremen.de>]

I have appended two datasets. One consists of survey data and the other 
one has information on stories published by a cable news channel. They 
have in common the variables DATE (expressed as, for instance 20080101) 
and, correspondingly, day (a numeric expression of DATE starting with 1 
for January 1st). For the survey data, these variables denote the date 
of the interview. For the media data, they denote the date a given story 
was aired.

What I would like to do is to assign each observation of the survey 
dataset the cumulated frequency of the media output of the 2 weeks prior 
to the interview date. For instance, if an interview was conducted on 
day 23, I would like to create a new variable stating how many stories 
were aired about a given issue (e.g. unemployment) on days 8 through 22.

I have read around quiet a bit and I imagine the solution will have 
something to do with running loops, but unfortunately I am too much of a 
Stata novice to figure this out by myself. Any help is greatly appreciated!
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/